Multi-Band Texture Image Fusion Based on the Embedded Multi-Scale Decomposition and Possibility Theory

Area of Science:

• Computer Vision
• Image Processing
• Remote Sensing

Background:

• Traditional dual-band image fusion methods struggle with multi-band images, especially those with texture features, leading to information loss.
• Sequential weighted averaging in existing techniques weakens inherent information, impacting target recognition and scene understanding.

Purpose of the Study:

• To propose a new multi-band texture image fusion method for more comprehensive and precise scene description.
• To address the information weakening issue in traditional multi-band image fusion.

Main Methods:

• A novel method combining embedded multi-scale decomposition and possibility theory for multi-band texture image fusion.
• Decomposition into high- and low-frequency components, with high-frequency fusion via maximum gray value and low-frequency fusion using a block-based approach with possibility theory.
• Integration of pixel-level and feature-level fusion, and space and transform domain technologies.

Main Results:

• The proposed method effectively fuses multi-band texture images, preserving inherent information.
• It successfully addresses the sawtooth effect at target edges through varied fusion rules for different block sizes.
• The method demonstrates improved scene description capabilities.

Conclusions:

• The developed multi-band texture image fusion method offers a significant improvement over existing techniques.
• It provides a more comprehensive and precise representation of scenes, beneficial for subsequent analysis.
• The integration of diverse techniques enhances robustness and accuracy in image fusion.